← Back to tool

What "AI Info" Means on a Facebook Post or Ad

Guide · updated 2 October 2026
Quick answer

Facebook's "AI info" label — the same system used on Instagram and Threads — fires when Meta reads a C2PA or IPTC provenance signal embedded in your image or video file, or when Meta's own generative tools were used to create it. In Ads Manager specifically, Meta has made disclosure mandatory: using Background Generation, Image Generation, or Add Animation inside the ad tools now applies the label automatically, with no advertiser input required.

Two different contexts, two slightly different triggers

Organic posts

For a regular post, Facebook reads the same embedded metadata signals as Instagram: C2PA Content Credentials and IPTC AI-source fields written by the tool that made or edited the image. If either is present, the post gets an "AI info" tag that viewers can tap for more context.

Ads

Facebook and Instagram ads are stricter. As of 2026, using any of Meta's built-in generative ad features (Background Generation, Image Generation, Add Animation) triggers the label automatically — Meta flags its own tool use directly, separate from reading file metadata. A first violation of the disclosure requirement can get an ad pulled and add a policy strike to the account, so advertisers using third-party AI tools for ad creative should assume the same metadata-reading logic as organic posts applies.

Why this matters for sellers and marketers specifically

If you're running product ads using AI-enhanced photography — background replacement, AI upscaling, generative fill to remove a distracting element — the resulting file may carry metadata that triggers disclosure even though the product itself is real and unedited in substance. Understanding exactly what's embedded, rather than guessing, matters more in an ads context where a labeling mistake carries an account-level penalty.

Checking what's actually in your file

Rather than assuming a file will or won't get flagged, AiEraser's View mode decodes the real C2PA manifest — claim generator name, recorded actions (created/converted/edited), and signature — so you know precisely what Meta's systems would read before you publish.

Frequently asked questions

Is the ad disclosure requirement different from the organic post label?
Yes — ads built with Meta's own in-tool generative features get labeled automatically regardless of metadata, while organic posts and ads using third-party tools are flagged based on the embedded C2PA/IPTC signal.
What happens on a first disclosure violation in Ads Manager?
Publicly described consequences include the ad being taken down and a policy strike added to the ad account — repeated violations typically escalate account-level restrictions.
Does Facebook's label work the same on Threads?
Threads uses the same underlying "AI info" labeling system as Facebook and Instagram, since all three read the same Meta-wide detection pipeline.

Related

Sources: Meta Transparency Center, Meta Business Help Center ad-disclosure documentation (2026). Platform ad policies change; verify current rules directly in Meta Ads Manager before relying on this for compliance purposes.